Previous Time Notepad content

2016-12-11 09:09:22

14:40 2016/10/30 Optical Character Recognition OCR
Radial basis function

Facial feature extraction with weighted local two-value pattern

Texture classification feature extraction based on local two-value pattern

Hello, thank you for your interest in our research work. The Brint code is not well-organized, but we have a newer approach, MRELBP, which can be found on the Oulu University MVG Group website. In their official website can also find some LBP method code, can be used for learning.
In addition, we have some other papers, I sent to you for reference.

A survey of local invariant features
Http://blog.sina.cn/dpool/blog/s/blog_52510b1d01014xks.html

1. Open the Software
2. Modify Parameters: direction0,1,2,3
Sensitivity parameter: 0-9
3.result error classification, calculation of false alarm rate

Cloud Habitat Conference Video

RMB direction Category:
1. Remove the raw file;
2. According to Grandpa Mao in top or bottom face, category 0-3; (by category)
3. Each category is paired and renamed according to Top->up,bottom->down rules

------------------------------------built Sample library algorithm ——————————————————————

Process:
Mian ()-->imagerecognition ()

-->imagevaluecheck ()-->locatemoneycorner () (Corner correction module-corner structure: contains four corner coordinates, four edges, etc.).

-->findcornerandangle ()-->getmoneyedgepoint ()--->

------------------------------------built Sample library algorithm end ——————————————————————

2016/10/24

1) Sample Finishing
2) Image processing program jobs
3) CMOS Commissioning
4) English
5) Correction Algorithm understanding

-------------------------AK Algorithm Flow----------------
1. Parse File path Pathanalyze
2. Load image data Loadbitmapfromfile: Iterate through the saved file name; The new code traverses the 0~n file from while (1)
3. Denomination identification Imagevaluecheck---->locateidcornernew--->findcornerandangle

Special samples for some judgments:

1. Two sides are not parallel, the difference is more than 8 degrees, then the confidence level is less than 40%, need to correct the sample.

2. Image imaging? Up/down

Opencv_core2411d.lib
Opencv_calib3d2411d.lib
Opencv_contrib2411d.lib
Opencv_features2d2411d.lib
Opencv_flann2411d.lib
Opencv_gpu2411d.lib
Opencv_highgui2411d.lib
Opencv_imgproc2411d.lib
Opencv_legacy2411d.lib
Opencv_ml2411d.lib
Opencv_objdetect2411d.lib
Opencv_ts2411d.lib
Opencv_video2411d.lib

Weekend Quests:

1. Academic Ethics Discussion
2. Lab Safety Exam
3. English Composition
4. The task of clearing the extension
5. Other Learning

Some machines may not have a code tray value

I'll make good use of GitHub, keep the idea of open source,
Positive contributions to my code! I often download open source code,
Learn some open source software. This is my GitHub Address:https://github.com/ranjiewwen.
From this can get my contibutions last years.
Now I am a graduate student, the direction of digital image processing and machine vision.
So I often contact with so many open source libraries, such as Opencv,opensift, Qt-creator, Python,linux,caffe,docker,sci Kit-learn,tensorflow and so on.
I apply to GitHub education project to improve myself.
If I get this chance, I'll cherish to use, learn more technology!

Previous Time Notepad content